The Big Sleep

This blog usually focuses on diet, because that's my specialty. But if you want Whole Health, you need the whole package: a diet and lifestyle that is broadly consistent with our evolutionary heritage. I think we all know that on some level, but a recent paper has reminded me of it.

I somehow managed to get on the press list of the Annals of Internal Medicine. That means they send me embargoed papers before they're released to the general public. That journal publishes a lot of high-impact diet studies, so it's a great privilege for me. I get to write about the studies, and publish my analysis at the time of general release, which is the same time the news outlets publish their stories.

One of the papers they sent me recently is a fat loss trial with an interesting twist (1; see below). All participants were told to eat 10% fewer calories that usual for two weeks, however half of them were instructed to sleep for 8 and a half hours per night, and the other half were instructed to sleep for 5 and a half hours*. The actual recorded sleep times were 7:25 and 5:14, respectively.

Weight loss by calorie restriction causes a reduction of both fat and lean mass, which is what the investigators observed. Both groups lost the same amount of weight. However, 80% of the weight was lost as fat in the high-sleep group (2.4/3.0 kg lost as fat), while only 48% of it was lost as fat in the low-sleep group (1.4/2.9 kg lost as fat). Basically, the sleep-deprived group lost as much lean mass as they did fat mass, which is not good!

There are many observational studies showing associations between insufficient sleep, obesity and diabetes. However, I think studies like that are particularly vulnerable to confounding variables, so I've never known quite what to make of them. Furthermore, they often show that long sleep duration associates with poor health as well, which I find highly unlikely to reflect cause and effect. I discussed one of those studies in a post a couple of years ago (2). That's why I appreciate this controlled trial so much.

Another sleep restriction trial published in the Lancet in 1999 showed that restricting healthy young men to four hours of sleep per night caused them to temporarily develop glucose intolerance, or pre-diabetes (3).

Furthermore, their daily rhythm of the hormone cortisol became abnormal. Rather than the normal pattern of a peak in the morning and a dip in the evening, sleep deprivation blunted their morning cortisol level and enhanced it in the evening. Cortisol is a stress hormone, among other things, and its fluctuations may contribute to our ability to feel awake in the morning and ready for bed at night.

The term "adrenal fatigue", which refers to the aforementioned disturbance in cortisol rhythm, is characterized by general fatigue, difficulty waking up in the morning, and difficulty going to sleep at night. It's a term that's commonly used by alternative medical practitioners but not generally accepted by mainstream medicine, possibly because it's difficult to demonstrate and the symptoms are fairly general. Robb Wolf talks about it in his book The Paleo Solution.

The investigators concluded:

Sleep debt has a harmful impact on carbohydrate metabolism and endocrine function. The effects are similar to those seen in normal ageing and, therefore, sleep debt may increase the severity of age-related chronic disorders.

So there you have it. Besides making us miserable, lack of sleep appears to predispose to obesity and diabetes, and probably sets us up for the Big Sleep down the line. I can't say I'm surprised, given how awful I feel after even one night of six hour sleep. I feel best after 9 hours, and I probably average about 8.5. Does it cut into my free time? Sure. But it's worth it to me, because it allows me to enjoy my day much more.

Keep your room as dark as possible during sleep. It also helps to avoid bright light, particularly in the blue spectrum, before bed (4). "Soft white" bulbs are preferable to full spectrum in the evening. If you need to use your computer, dim the monitor and adjust it to favor warm over cool colors. For people who sleep poorly due to anxiety, meditation before bed can be highly effective. I posted a tutorial here.

1. Nedeltcheva, AV et al. "Insufficient Sleep Undermines Dietary Efforts to Reduce Adiposity." Annals of Internal Medicine. 2010. Advanced publication.


* The study was a randomized crossover design with a 3 month washout period, which I consider a rigorous design. I think the study overall was very clever. The investigators used calorie restriction to cause rapid changes in body composition so that they could see differences on a reasonable timescale, rather than trying to deprive people of sleep for months and look for more gradual body fat changes without dietary changes. The latter experiment would have been more interesting, but potentially impractical and unethical.

Potatoes and Human Health, Part III

Potato-eating Cultures: the Quechua

The potato is thought to have originated in what is now Peru, on the shores of lake Titicaca. Native Peruvians such as the Quechua have been highly dependent on the potato for thousands of years. A 1964 study of the Quechua inhabitants of Nuñoa showed that they obtained 74% of their calories from potatoes (fresh and chuños), 10% from grains, 10% from Chenopodia (quinoa and cañihua), and 4% from animal foods. Total energy intake was 3,170 calories per day (1).

In 2001, a medical study of rural Quechua men reported an average body fat percentage of 16.4% (2). The mean age of the volunteers was 38. Body fat did increase slowly with age in this population, and by age 65 it was predicted to be about 20% on average. That's below the threshold of overweight, so I conclude that most men in this population are fairly lean, although there were a few overweight individuals.

In 2004, a study in rural Quechua women reported a body fat percentage of 31.2% in volunteers with a mean age of 35 (3). Body fat percentage was higher in a group of Quechua immigrants to the Peruvian capital of Lima. Among rural women, average fasting insulin was 6.8 uIU/mL, and fasting glucose was 68.4 mg/dL, which together suggest good insulin sensitivity and glucose control (4). Insulin and glucose were considerably lower in the rural group than the urban group. Blood pressure was low in both groups. Overall, this suggests that overweight is common among Quechua women.

Rural Quechua are characteristically short, with the average adult man standing no more than 5' 2" (2). One might be tempted to speculate that this reflects stunting due to a deficient diet. However, given the fact that nearly all non-industrial populations, including contemporary hunter-gatherers, are short by modern standards, I'm not convinced the Quechua are abnormal. A more likely explanation is that industrial foods cause excessive tissue growth in modern populations, perhaps by promoting overeating and excessive insulin and IGF-1 production, which are growth factors. I first encountered this hypothesis in Dr. Staffan Lindeberg's book Food and Western Disease.

I don't consider the Quechua diet to be optimal, but it does seem to support a reasonable level of metabolic health. Rural Quechua men subsisting on potatoes are relatively lean, while women are often overweight, though less overweight than urban Quechua who eat fewer potatoes. Unfortunately, I don't have more detailed data on other aspects of their health, such as gastrointestinal health.

Potato-eating Cultures: the Aymara


The Aymara are another potato-dependent people of the Andes, who span Peru, Bolivia and Chile. The first paper I'll discuss is titled "Low Prevalence of Type II Diabetes Despite a High Body Mass Index in the Aymara Natives From Chile", by Dr. Jose Luis Santos and colleagues (5). In the paper, they show that the prevalence of diabetes in this population was 1.5%, and the prevalence of pre-diabetes was 3.6%. The prevalence of both remained low even in the elderly. Here's a comparison of those numbers with figures from the modern United States (6):

That's quite a difference! The prevalence of diabetes in this population is low, but not as low as in some cultures such as the Kitavans (7, 8).

Now to discuss the "high body mass index" referenced in the title of the paper. The body mass index (BMI) is the relation between height and weight, and often, but not always, reflects fatness. The average BMI of this population was 24.9, which is very close to the cutoff between normal and overweight (25).

Investigators were surprised to find such a low prevalence of diabetes in this population, despite their apparent high prevalence of overweight. Yet if you've seen pictures of rural native South Americans, you may have noticed they're often built short and thick, with wide hips and big barrel chests. Could this be confounding the relationship between BMI and body fatness? To answer that question, I found another paper that estimated Aymara body fat using skinfold measurements (9). That study reported that both men and women remained relatively lean throughout life (ages 4-65), with only two of 23 subjects classified as overweight on the basis of body fat percentage, and none classified as obese.

Back to the first paper. In this Aymara group, blood pressure was on the high side. Serum cholesterol was also a bit high for a traditionally-living population, but still lower than most modern groups (~188 mg/dL). I find it very interesting that the cholesterol level in this population that eats virtually no fat was the same as on Tokelau, where nearly half of calories come from highly saturated coconut fat (10, 11). Fasting insulin is also on the high side in the Aymara, which is also interesting given their good glucose tolerance and low prevalence of diabetes.  This could be related to differences in the measurement assay rather than a true difference in fasting insulin, since these assays were often not standardized between studies.

Together, this shows that a lifetime of high-carbohydrate, high-glycemic food does not necessarily lead to overweight or metabolic problems in the context of a traditional diet and lifestyle.

Potato-eating Cultures: the Irish

Potatoes were introduced to Ireland in the 17th century. They were well suited to the cool, temperate climate, and more productive than any other crop. By the early 18th century, potatoes were the main source of calories, particularly for the poor who ate practically nothing else. In 1839, the average Irish laborer obtained 87% of his calories from potatoes (12). In 1845, the potato blight Phytophthora infestans struck, decimating potato plantations nationwide and creating the Great Famine.

There isn't much reliable information on the health status of the Irish prior to the famine, besides reports of vitamin A deficiency symptoms (13) due to the fact that neither of the primary articles of diet, potatoes and buttermilk, provide significant vitamin A. However, the Irish at the time had a very high fertility rate, and anecdotal reports described them as healthy and attractive (14):

As far as fecundity is concerned, the high nutritional value of the potato diet might have played a significant role, but little supportive evidence has been presented so far... What is known is that the Irish in general and Irish women in particular were widely described as healthy and good-looking. Adam Smith's famous remark that potatoes were "peculiarly suitable to the health of the human constitution" can be complemented with numerous observations from other contemporary observers to the same effect.

Controlled Feeding Studies

Starting nearly a century ago, a few researchers decided to feed volunteers potato-only diets to achieve various research objectives. The first such experiment was carried out by a Dr. M. Hindhede and published in 1913 (described in 15). Hindhede's goal was to explore the lower limit of the human protein requirement and the biological quality of potato protein. He fed three healthy adult men almost nothing but potatoes and margarine for 309 days (margarine was not made from hydrogenated seed oils at the time), all while making them do progressively more demanding physical labor. They apparently remained in good physical condition. Here's a description of one of his volunteers, a Mr. Madsen, from another book (described in 16; thanks to Matt Metzgar):

In order to test whether it was possible to perform heavy work on a strict potato diet, Mr. Madsen took a place as a farm laborer... His physical condition was excellent. In his book, Dr. Hindhede shows a photograph of Mr. Madsen taken on December 21st, 1912, after he had lived for almost a year entirely on potatoes. This photograph shows a strong, solid, athletic-looking figure, all of whose muscles are well-developed, and without excess fat. ...Hindhede had him examined by five physicians, including a diagnostician, a specialist in gastric and intestinal diseases, an X-ray specialist, and a blood specialist. They all pronounced him to be in a state of perfect health.

Dr. Hindhede discovered that potato protein is high quality, providing all essential amino acids and high digestibility. Potato protein alone is sufficient to sustain an athletic man (although that doesn't make it optimal). A subsequent potato feeding study published in 1927 confirmed this finding (17). Two volunteers, a man and a woman, ate almost nothing but potatoes with a bit of lard and butter for 5.5 months. The man was an athlete but the woman was sedentary. Body weight and nitrogen balance (reflecting protein gain/loss from the body) remained constant throughout the experiment, indicating that their muscles were not atrophying at any appreciable rate, and they were probably not putting on fat. The investigators remarked:

The digestion was excellent throughout the experiment and both subjects felt very well. They did not tire of the uniform potato diet and there was no craving for change.

In one of his Paleo Diet newsletters titled "Consumption of Nightshade Plants (Part 1)", Dr. Loren Cordain referenced two feeding studies showing that potatoes increase the serum level of the inflammatory cytokine interleukin-6 (22, 23). However, one study was not designed to determine the specific role of potato in the change (two dietary factors were altered simultaneously), and the other used potato chips as the source of potato. So I don't find these studies particularly relevant to the question at hand.

Just yesterday, Chris Voigt of the Washington State Potato Commission embarked on his own n=1 potato feeding experiment as a way to promote Washington state potatoes. He'll be eating nothing but potatoes and a little fat for two months, and getting a full physical at the end. Check out his website for more information and updates (18). Mr. Voigt has graciously agreed to a written interview with Whole Health Source at the end of his experiment. He pointed out to me that the Russet Burbank potato, the most popular variety in the United States, is over 135 years old. Stay tuned for more interesting facts from Mr. Voigt in early December.

Observational Studies

With the recent interest in the health effects of the glycemic index, a few studies have examined the association between potatoes and health in various populations. The results are all over the place, with some showing positive associations with health, and others showing negative associations (19, 20, 21). As a whole, I find these studies difficult to interpret and not very helpful.

Anecdotes

Some people feel good when they eat potatoes. Others find that potatoes and other members of the nightshade family give them digestive problems, exacerbate their arthritis, or cause fat gain. I haven't encountered any scientific research to substantiate claims that nightshades aggravate arthritis or other inflammatory conditions. However, that doesn't necessarily mean there aren't individuals who are sensitive. If potatoes don't agree with you, by all means avoid them.

The Bottom Line

You made it to the end! Give yourself a pat on the back. You deserve it.

In my opinion, the scientific literature as a whole, including animal and human studies, suggests rather consistently that potatoes can be a healthy part of a varied diet for most people, and they probably do not generally promote digestive problems, fat gain, or metabolic dysfunction.  Nevertheless, I wouldn't recommend eating nothing but potatoes for any length of time. If you do choose to eat potatoes, follow these simple guidelines:

• Don't eat potatoes that are green, sprouting, blemished, or damaged
• Store them in a cool, dark place. They don't need to be refrigerated but it will extend their life
  
• Peel them before eating if you rely on them as a staple food
  

Enjoy your potatoes!

Income, obesity, and heart disease in US states

The figure below combines data on median income by state (bottom-left and top-right), as well as a plot of heart disease death rates against percentage of population with body mass index (BMI) greater than 30 percent. The data are recent, and have been provided by CNN.com and creativeclass.com, respectively.

Heart disease deaths and obesity are strongly associated with each other, and both are inversely associated with median income. US states with lower median income tend to have generally higher rates of obesity and heart disease deaths.

The reasons are probably many, complex, and closely interconnected. Low income is usually associated with high rates of stress, depression, smoking, alcoholism, and poor nutrition. Compounding the problem, these are normally associated with consumption of cheap, addictive, highly refined foods.

Interestingly, this is primarily an urban phenomenon. If you were to use hunter-gatherers as your data sources, you would probably see the opposite relationship. For example, non-westernized hunter-gatherers have no income (at least not in the “normal” sense), but typically have a lower incidence of obesity and heart disease than mildly westernized ones. The latter have some income.

Tragically, the first few generations of fully westernized hunter-gatherers usually find themselves in the worst possible spot.

Potatoes and Human Health, Part II

Glycoalkaloids in Commonly Eaten Potatoes

Like many edible plants, potatoes contain substances designed to protect them from marauding creatures. The main two substances we're concerned with are alpha-solanine and alpha-chaconine, because they are the most toxic and abundant. Here is a graph of the combined concentration of these two glycoalkaloids in common potato varieties (1):

We can immediately determine three things from this graph:

• Different varieties contain different amounts of glycoalkaloids.
• Common commercial varieties such as russet and white potatoes are low in glycoalkaloids. This is no accident. The glycoalkaloid content of potatoes is monitored in the US.
  
• Most of the glycoalkaloid content is in the skin (within 1 mm of the surface). That way, predators have to eat through poison to get to the flesh. Fortunately, humans have peelers.

I'll jump the gun and tell you that the generally accepted safe level of potato glycoalkaloids is 200 mcg/g fresh weight (1). You can see that all but one variety are well below this level when peeled. Personally, I've never seen the Snowden variety in the store or at the farmer's market. It appears to be used mostly for potato chips.

Glycoalkaloid Toxicity in Animals

Potato glycoalkaloids are undoubtedly toxic at high doses. They have caused many harmful effects in animals and humans, including (1, 2):

• Death (humans and animals)
  
• Weight loss, diarrhea (humans and animals)
  
• Anemia (rabbits)
• Liver damage (rats)
  
• Lower birth weight (mice)
  
• Birth defects (in animals injected with glycoalkaloids)
  
• Increased intestinal permeability (mice)

However, it's important to remember the old saying "the dose makes the poison". The human body is designed to handle a certain amount of plant toxins with no ill effects. Virtually every plant food, and a few animal foods, contains some kind of toxic substance. We're constantly bombarded by gamma rays, ultra violet rays, bacterial toxins, free radicals, and many other potentially harmful substances. In excess, they can be deadly, but we are adapted to dealing with small amounts of them, and the right dose can even be beneficial in some cases.

All of the studies I mentioned above, except one, involved doses of glycoalkaloids that exceed what one could get from eating typical potatoes. They used green or blemished potatoes, isolated potato skins, potato sprouts or isolated glycoalkaloids (more on this later). The single exception is the last study, showing that normal doses of glycoalkaloids can aggravate inflammatory bowel disease in transgenic mice that are genetically predisposed to it (3)*.

What happens when you feed normal animals normal potatoes? Not much. Many studies have shown that they suffer no ill effects whatsoever, even at high intakes (1, 2). This has been shown in primates as well (4, 5, 6). In fact, potato-based diets appear to be generally superior to grain-based diets in animal feed. As early as 1938, Dr. Edward Mellanby showed that grains, but not potatoes, aggravate vitamin A deficiency in rats and dogs (7). This followed his research showing that whole grains, but not potatoes, aggravate vitamin D deficiency due to their high phytic acid content (Mellanby. Nutrition and Disease. 1934). Potatoes were also a prominent part of Mellanby's highly effective tooth decay reversal studies in humans, published in the British Medical Journal in 1932 (8, 9).

Potatoes partially protect rats against the harmful effects of excessive cholesterol feeding, when compared to wheat starch-based feed (10). Potato feeding leads to a better lipid profile and intestinal short-chain fatty acid production than wheat starch or sugar in rats (11). I wasn't able to find a single study showing any adverse effect of normal potato feeding in any normal animal. That's despite reading two long review articles on potato glycoalkaloids and specifically searching PubMed for studies showing a harmful effect. If you know of one, please post it in the comments section.

In the next post, I'll write about the effects of potatoes in the human diet, including data on the health of traditional potato-eating cultures... and a curious experiment by the Washington State Potato Commission that will begin on October 1.


*Interleukin-10 knockout mice. IL-10 is a cytokine involved in the resolution of inflammation and these mice develop inflammatory bowel disease (regardless of diet) due to a reduced capacity to resolve inflammation.

Low nonexercise activity thermogenesis: Uncooperative genes or comfy furniture?

The degree of nonexercise activity thermogenesis (NEAT) seems to a major factor influencing the amount of fat gained or lost by an individual. It also seems to be strongly influenced by genetics, because NEAT is largely due to involuntary activities like fidgeting.

But why should this be?

The degree to which different individuals will develop diseases of civilization in response to consumption of refined carbohydrate-rich foods can also be seen as influenced by genetics. After all, there are many people who eat those foods and are thin and healthy, and that appears to be in part a family trait. But whether we consume those products or not is largely within our control.

So, it is quite possible that NEAT is influenced by genetics, but the fact that NEAT is low in so many people should be a red flag. In the same way that the fact that so many people who eat refined carbohydrate-rich foods are obese should be a red flag. Moreover, modern isolated hunter-gatherers tend to have low levels of body fat. Given the importance of NEAT for body fat regulation, it is not unreasonable to assume that NEAT is elevated in hunter-gatherers, compared to modern urbanites. Hunter-gatherers live more like our Paleolithic ancestors than modern urbanites.

True genetic diseases, caused by recent harmful mutations, are usually rare. If low NEAT were truly a genetic “disease”, those with low NEAT should be a small minority. That is not the case. It is more likely that the low NEAT that we see in modern urbanites is due to a maladaptation of our Stone Age body to modern life, in the same way that our Stone Age body is maladapted to the consumption of foods rich in refined grains and seeds.

What could have increased NEAT among our Paleolithic ancestors, and among modern isolated hunter-gatherers?

One thing that comes to mind is lack of comfortable furniture, particularly comfortable chairs (photo below from: prlog.org). It is quite possible that our Paleolithic ancestors invented some rudimentary forms of furniture, but they would have been much less comfortable than modern furniture used in most offices and homes. The padding of comfy office chairs is not very easy to replicate with stones, leaves, wood, or even animal hides. You need engineering to design it; you need industry to produce that kind of thing.

I have been doing a little experiment with myself, where I do things that force me to sit tall and stand while working in my office, instead of sitting back and “relaxing”. Things like putting a pillow on the chair so that I cannot rest my back on it, or placing my computer on an elevated surface so that I am forced to work while standing up. I tend to move a lot more when I do those things, and the movement is largely involuntary. These are small but constant movements, a bit like fidgeting. (It would be interesting to tape myself and actually quantify the amount of movement.)

It seems that one can induce an increase in NEAT, which is largely due to involuntary activities, by doing some voluntary things like placing a pillow on a chair or working while standing up.

Is it possible that the unnaturalness of comfy furniture, and particularly of comfy chairs, is contributing (together with other factors) to not only making us fat but also having low-back problems?

Both obesity and low-back problems are widespread among modern urbanites. Yet, from an evolutionary perspective, they should not be. They likely impaired survival success among our ancestors, and thus impaired their reproductive success. Evolution “gets angry” at these things; over time it wipes them out. In my reading of studies of hunter-gatherers, I don’t recall a single instance in which obesity and low-back problems were described as being widespread.

Potatoes and Human Health, Part I

Potatoes: an Introduction

Over 10,000 years ago, on the shores of lake Titicaca in what is now Peru, a culture began to cultivate a species of wild potato, Solanum tuberosum. They gradually transformed it into a plant that efficiently produces roundish starchy tubers, in a variety of strains that suited the climactic and gastronomic needs of various populations. These early farmers could not have understood at the time that the plant they were selecting would become the most productive crop in the world*, and eventually feed billions of people around the globe.

Wild potatoes, which were likely consumed by hunter-gatherers before domestication, are higher in toxic glycoalkaloids. These are defensive compounds that protect against insects, infections and... hungry animals. Early farmers selected varieties that are low in bitter glycoalkaloids, which are the ancestors of most modern potatoes, however they didn't abandon the high-glycoalkaloid varieties. These were hardier and more tolerant of high altitudes, cold temperatures and pests. Cultures living high in the Andes developed a method to take advantage of these hardy but toxic potatoes, as well as their own harsh climate: they invented chuños. These are made by leaving potatoes out in the open, where they are frozen at night, stomped underfoot and dried in the sun for many days**. What results is a dried potato with a low glycoalkaloid content that can be stored for a year or more.

Nutritional Qualities

From a nutritional standpoint, potatoes have a bad reputation, but this is undeserved in my opinion. If I had to pick a single food to eat exclusively for an extended period of time, potatoes would be high on the list. One reason is that they contain an adequate amount of complete protein, meaning they don't have to be mixed with another protein source as with grains and legumes. Another reason is that a number of cultures throughout history have successfully relied on the potato as their principal source of calories, and several continue to do so. A third reason is that they're eaten in an unrefined, fresh state.

Potatoes contain an adequate amount of many essential minerals, and due to their low phytic acid content (1), the minerals they contain are well absorbed. They're rich in magnesium and copper, two minerals that are important for insulin sensitivity and cardiovascular health (2, 3). They're also high in potassium, which helps control blood pressure, and vitamin C. Overall, they have a micronutrient content that compares favorably with other starchy root vegetables such as taro and cassava (4, 5, 6), and they offer considerably more micronutrients than refined carbohydrates such as white flour, white rice and white sugar.

On the other hand, I don't have to eat potatoes exclusively, so what do they have to offer a mixed diet? They have a high glycemic index, which means they raise blood sugar more than an equivalent serving of most carbohydrate foods, although I'm not convinced that's a problem in people with good blood sugar control (7, 8). They contain adequate fiber, but less than some other sources of starch. For example, sweet potatoes, an unrelated species, contain more micronutrients and fiber, and have been a central food source for healthy cultures (9). However, the main reasons temperate-climate cultures throughout the world eat potatoes is they yield well, they're easily digested, they fill you up and they taste good.

In the next post, I'll delve into the biology and toxicology of potato glycoalkaloids, and review some animal data. In further posts, I'll address the most important question of all: what happens when a person eats mostly potatoes... for months, years, and generations?


* In terms of calories produced per acre.

** A simplified description. The process can actually be rather involved, with several different drying, stomping and leaching steps.

Strong causation can exist without any correlation: The strange case of the chain smokers, and a note about diet

Researchers like to study samples of data and look for associations between variables. Often those associations are represented in the form of correlation coefficients, which go from -1 to 1. Another popular measure of association is the path coefficient, which usually has a narrower range of variation. What many researchers seem to forget is that the associations they find depend heavily on the sample they are looking at, and on the ranges of variation of the variables being analyzed.

A forgotten warning: Causation without correlation

Often those who conduct multivariate statistical analyses on data are unaware of certain limitations. Many times this is due to lack of familiarity with statistical tests. One warning we do see a lot though is: Correlation does not imply causation. This is, of course, absolutely true. If you take my weight from 1 to 20 years of age, and the price of gasoline in the US during that period, you will find that they are highly correlated. But common sense tells me that there is no causation whatsoever between these two variables.

So correlation does not imply causation alright, but there is another warning that is rarely seen: There can be strong causation without any correlation. Of course this can lead to even more bizarre conclusions than the “correlation does not imply causation” problem. If there is strong causation between variables B and Y, and it is not showing as a correlation, another variable A may “jump in” and “steal” that “unused correlation”; so to speak.

The chain smokers “study”

To illustrate this point, let us consider the following fictitious case, a study of “100 cities”. The study focuses on the effect of smoking and genes on lung cancer mortality. Smoking significantly increases the chances of dying from lung cancer; it is a very strong causative factor. Here are a few more details. Between 35 and 40 percent of the population are chain smokers. And there is a genotype (a set of genes), found in a small percentage of the population (around 7 percent), which is protective against lung cancer. All of those who are chain smokers die from lung cancer unless they die from other causes (e.g., accidents). Dying from other causes is a lot more common among those who have the protective genotype.

(I created this fictitious data with these associations in mind, using equations. I also added uncorrelated error into the equations, to make the data look a bit more realistic. For example, random deaths occurring early in life would reduce slightly any numeric association between chain smoking and cancer deaths in the sample of 100 cities.)

The table below shows part of the data, and gives an idea of the distribution of percentage of smokers (Smokers), percentage with the protective genotype (Pgenotype), and percentage of lung cancer deaths (MLCancer). (Click on it to enlarge. Use the "CRTL" and "+" keys to zoom in, and CRTL" and "-" to zoom out.) Each row corresponds to a city. The rest of the data, up to row 100, has a similar distribution.

The graphs below show the distribution of lung cancer deaths against: (a) the percentage of smokers, at the top; and (b) the percentage with the protective genotype, at the bottom. Correlations are shown at the top of each graph. (They can vary from -1 to 1. The closer they are to -1 or 1, the stronger is the association, negative or positive, between the variables.) The correlation between lung cancer deaths and percentage of smokers is slightly negative and statistically insignificant (-0.087). The correlation between lung cancer deaths and percentage with the protective genotype is negative, strong, and statistically significant (-0.613).

Even though smoking significantly increases the chances of dying from lung cancer, the correlations tell us otherwise. The correlations tell us that lung cancer does not seem to cause lung cancer deaths, and that having the protective genotype seems to significantly decrease cancer deaths. Why?

If there is no variation, there is no correlation

The reason is that the “researchers” collected data only about chain smokers. That is, the variable “Smokers” includes only chain smokers. If this was not a fictitious case, focusing the study on chain smokers could be seen as a clever strategy employed by researchers funded by tobacco companies. The researchers could say something like this: “We focused our analysis on those most likely to develop lung cancer.” Or, this could have been the result of plain stupidity when designing the research project.

By restricting their study to chain smokers the researchers dramatically reduced the variability in one particular variable: the extent to which the study participants smoked. Without variation, there can be no correlation. No matter what statistical test or software is used, no significant association will be found between lung cancer deaths and percentage of smokers based on this dataset. No matter what statistical test or software is used, a significant and strong association will be found between lung cancer deaths and percentage with the protective genotype.

Of course, this could lead to a very misleading conclusion. Smoking does not cause lung cancer; the real cause is genetic.

A note about diet

Consider the analogy between smoking and consumption of a particular food, and you will probably see what this means for the analysis of observational data regarding dietary choices and disease. This applies to almost any observational study, including the China Study. (Studies employing experimental control manipulations would presumably ensure enough variation in the variables studied.) In the China Study, data from dozens of counties were collected. One may find a significant association between consumption of food A and disease Y.

There may be a much stronger association between food B and disease Y, but that association may not show up in statistical analyses at all, simply because there is little variation in the data regarding consumption of food B. For example, all those sampled may have eaten food B; about the same amount. Or none. Or somewhere in between, within a rather small range of variation.

Statistical illiteracy, bad choices, and taxation

Statistics is a “necessary evil”. It is useful to go from small samples to large ones when we study any possible causal association. By doing so, one can find out whether an observed effect really applies to a larger percentage of the population, or is actually restricted to a small group of individuals. The problem is that we humans are very bad at inferring actual associations from simply looking at large tables with numbers. We need statistical tests for that.

However, ignorance about basic statistical phenomena, such as the one described here, can be costly. A group of people may eliminate food A from their diet based on coefficients of association resulting from what seem to be very clever analyses, replacing it with food B. The problem is that food B may be equally harmful, or even more harmful. And, that effect may not show up on statistical analyses unless they have enough variation in the consumption of food B.

Readers of this blog may wonder why we explicitly use terms like “suggests” when we refer to a relationship that is suggested by a significant coefficient of association (e.g., a linear correlation). This is why, among other reasons.

One does not have to be a mathematician to understand basic statistical concepts. And doing so can be very helpful in one’s life in general, not only in diet and lifestyle decisions. Even in simple choices, such as what to be on. We are always betting on something. For example, any investment is essentially a bet. Some outcomes are much more probable than others.

Once I had an interesting conversation with a high-level officer of a state government. I was part of a consulting team working on an information technology project. We were talking about the state lottery, which was a big source of revenue for the state, comparing it with state taxes. He told me something to this effect:

Our lottery is essentially a tax on the statistically illiterate.